beecoder-playground / README.md
pszemraj's picture
โœ๏ธ
cc59e08
---
title: BeeCoder Demo
emoji: ๐Ÿ
colorFrom: gray
colorTo: yellow
sdk: gradio
sdk_version: 3.28.3
app_file: app.py
pinned: true
license: apache-2.0
---
# ๐ŸBeeCoder Demo๐Ÿ
## Code-Completion Playground ๐Ÿ’ป with ๐Ÿ[BeeCoder](https://huggingface.co/BEE-spoke-data/smol_llama-101M-GQA-python) Models
This is a demo playground for generating Python code with the power of ๐Ÿ[BeeCoder](https://huggingface.co/BEE-spoke-data/smol_llama-101M-GQA-python), a **fine-tuned** version of the tiny [101M base model](https://huggingface.co/BEE-spoke-data/smol_llama-101M-GQA) on a dataset of pypi packages.
โ„น๏ธ This is not an instruction model but just a code completion tool.
---
**Intended Use**: This app and its [supporting model](https://huggingface.co/BEE-spoke-data/smol_llama-101M-GQA-python) are provided for demonstration purposes only; not to serve as a replacement for human expertise. For more details on the model, please refer to the [model card](https://huggingface.co/BEE-spoke-data/smol_llama-101M-GQA-python).
In our country, we say _"To let 100M parameters model generate python script and not validate is like to let monkey fly a plane"_. So please be careful with the generated code.
---
## Base Model Information
The base model, smol_llama-101M-GQA, has been pre-trained on a relatively small number of high quality tokens (less than ~20B). It has impressive performance despite its compact size of 101M parameters. Training data for this base model included:
- [JeanKaddour/minipile](https://huggingface.co/datasets/JeanKaddour/minipile)
- [pszemraj/simple_wikipedia_LM](https://huggingface.co/datasets/pszemraj/simple_wikipedia_LM)
- [BEE-spoke-data/wikipedia-20230901.en-deduped](https://huggingface.co/datasets/BEE-spoke-data/wikipedia-20230901.en-deduped)
- [mattymchen/refinedweb-3m](https://huggingface.co/datasets/mattymchen/refinedweb-3m)
You can find more information about the base model [here](https://huggingface.co/BEE-spoke-data/smol_llama-101M-GQA).
---
### Credits
This app is modified from a demo playground originally built for [StarCoder](https://huggingface.co/bigcode/starcoder) by [BigCode](https://huggingface.co/bigcode). You can find the original demo [here](https://huggingface.co/spaces/bigcode/bigcode-playground).
---